2017 SEAOC CONVENTION PROCEEDINGS 1 Seismically-Resilient Stair Systems for Buildings Cameron J. Black, Ph.D. P.E., Associate Ian D. Aiken, Ph.D. P.E., Principal Seismic Isolation Engineering, Inc. Berkeley, CA Kevin W. Smith, P.E. Robert J. Belvin, P.E. Anthony J. Peachey Construction Specialties, Inc. Muncy, PA Abstract There is an increasing understanding that structures and communities need to be resilient. For structures in seismically active areas resiliency is strongly correlated with how buildings perform during and following an earthquake. Safe egress for occupants and ingress for emergency personnel is a critical component of a building’s performance and is a matter of life safety during, and after a seismic event. For multistory buildings, stairs are essential for safe egress. Several recent earthquakes have highlighted the vulnerability of stairs and recent experimental testing has revealed that safe egress from buildings can be compromised even when the building drift demands are much lower than the design level performance target of the building. This paper presents engineered solutions for seismically resilient stairs that have been recently developed and tested. Full-scale shake table testing of several different stair systems has been conducted and the systems all shown to perform well. This paper discusses the results of testing and solutions that may be integrated to limit damage and provide safe building egress after significant earthquake events. 1. Introduction There is limited work in the literature which focuses specifically on the seismic performance of stairs. Roha et al. (1982) conducted an extensive review of stair damage sustained in earthquakes dating back to the 1906 San Francisco earthquake with accounts of “wrecked” stairways. The study presented both observed damage to stairs as well as damage of the primary structure which could be attributed to the stairs themselves. In the 1933 Long Beach earthquake, for example, it was observed that stairs acted like diagonal bracing between floors causing localized overloading of the structure. More recent accounts include Li and Mosalam (2013) who reported on failures of concrete stairs observed during the 2008 Wenchuan earthquake as well as numerous reports documenting damage to stairs in Christchurch during the 2010 Darfield and 2011 Lyttleton (Christchurch) earthquakes. A report to the New Zealand Royal Commission by Bull (2011) stated that the performance of stairs was less than desirable as stairs collapsed in at least four multi-story buildings and numerous others were seriously damaged. Beca (2011) presented the findings of a thorough investigation of the Forsyth Barr Building stair collapse which resulted from inadequate allowance for movement in the gap-and-ledge stair detail designed to allow the stairs to slide in order to accommodate movement associated with interstory drift. In response to concerns about stair collapses in the Christchurch central business district the New Zealand Department of Building and Housing issued Practice Advisory 13. The advisory focuses on existing buildings that had been designed with details similar to the Forsyth Barr Building. The advisory recommends that new buildings use details which allow stair flights to slide on landings without restriction and be capable of sustaining at least twice the Ultimate Limit State level inter- story displacements after allowing for construction tolerances. Recent experimental testing has also demonstrated the potential vulnerability of stairs in seismic events. Higgins (2009) performed quasi-static testing on two full-sized prefabricated steel stair assemblies with landings to assess their response when subjected to interstory drift in both directions while supporting gravity load. Although both assemblies completed the loading protocol up to 2.5% drift, it was concluded that the imposed lateral drift placed high deformation demands on the stair-to-landing connections and that the overall performance of the stairs is dependent on the
10
Embed
C Seismically-Resilient Stair Systems for · PDF fileSeismically-Resilient Stair Systems for Buildings ... to quantify the effect of stairs on the building ... studied the analytical
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
2017 SEAOC CONVENTION PROCEEDINGS
1
Seismically-Resilient Stair Systems for Buildings
Cameron J. Black, Ph.D. P.E., Associate Ian D. Aiken, Ph.D. P.E., Principal
Seismic Isolation Engineering, Inc. Berkeley, CA
Kevin W. Smith, P.E. Robert J. Belvin, P.E. Anthony J. Peachey
Construction Specialties, Inc. Muncy, PA
Abstract There is an increasing understanding that structures and
communities need to be resilient. For structures in seismically
active areas resiliency is strongly correlated with how
buildings perform during and following an earthquake. Safe
egress for occupants and ingress for emergency personnel is a
critical component of a building’s performance and is a matter
of life safety during, and after a seismic event.
For multistory buildings, stairs are essential for safe egress.
Several recent earthquakes have highlighted the vulnerability
of stairs and recent experimental testing has revealed that safe
egress from buildings can be compromised even when the
building drift demands are much lower than the design level
performance target of the building. This paper presents
engineered solutions for seismically resilient stairs that have
been recently developed and tested. Full-scale shake table
testing of several different stair systems has been conducted
and the systems all shown to perform well. This paper
discusses the results of testing and solutions that may be
integrated to limit damage and provide safe building egress
after significant earthquake events.
1. Introduction
There is limited work in the literature which focuses
specifically on the seismic performance of stairs. Roha et al.
(1982) conducted an extensive review of stair damage
sustained in earthquakes dating back to the 1906 San Francisco
earthquake with accounts of “wrecked” stairways. The study
presented both observed damage to stairs as well as damage of
the primary structure which could be attributed to the stairs
themselves. In the 1933 Long Beach earthquake, for example,
it was observed that stairs acted like diagonal bracing between
floors causing localized overloading of the structure. More
recent accounts include Li and Mosalam (2013) who reported
on failures of concrete stairs observed during the 2008
Wenchuan earthquake as well as numerous reports
documenting damage to stairs in Christchurch during the 2010
Darfield and 2011 Lyttleton (Christchurch) earthquakes. A
report to the New Zealand Royal Commission by Bull (2011)
stated that the performance of stairs was less than desirable as
stairs collapsed in at least four multi-story buildings and
numerous others were seriously damaged. Beca (2011)
presented the findings of a thorough investigation of the
Forsyth Barr Building stair collapse which resulted from
inadequate allowance for movement in the gap-and-ledge stair
detail designed to allow the stairs to slide in order to
accommodate movement associated with interstory drift. In
response to concerns about stair collapses in the Christchurch
central business district the New Zealand Department of
Building and Housing issued Practice Advisory 13. The
advisory focuses on existing buildings that had been designed
with details similar to the Forsyth Barr Building. The advisory
recommends that new buildings use details which allow stair
flights to slide on landings without restriction and be capable
of sustaining at least twice the Ultimate Limit State level inter-
story displacements after allowing for construction tolerances.
Recent experimental testing has also demonstrated the
potential vulnerability of stairs in seismic events. Higgins
(2009) performed quasi-static testing on two full-sized
prefabricated steel stair assemblies with landings to assess
their response when subjected to interstory drift in both
directions while supporting gravity load. Although both
assemblies completed the loading protocol up to 2.5% drift, it
was concluded that the imposed lateral drift placed high
deformation demands on the stair-to-landing connections and
that the overall performance of the stairs is dependent on the
2017 SEAOC CONVENTION PROCEEDINGS
2
deformability and ‘endurance’ of these connections. For that
reason, the paper recommends that careful detailing,
fabrication, and inspection of the stair-to-landing connections
be required to ensure desired system performance.
Shake table testing of a full-scale, five-story reinforced
concrete building was conducted at UCSD in 2013 with one
objective being to study the response of non-structural
components including piping, HVAC, sprinklers, a passenger
elevator and stairs (Chen et al. 2012). Wang et al. (2013)
summarize the extensive shake table test program and reports
that severe failure of the stair flight to floor slab connections
at multiple levels occurred during the design event. In Wang
et al. (2015), which focuses specifically on the response of the
stairs in the UCSD testing, it is reported that safe egress from
the building was compromised even when the associated drift
demands were much lower than the design performance target
of the building (with damage at 0.74% and 1.41%, compared
to the design target of 2.5%). Connection plate yielding was
observed at the landing and weld fractures were observed in
several locations with complete detachment observed at the
bottom of the 2nd floor landing (see Figure 1). A photo very
similar to Figure 1 d) which shows yielding at the connection
plate is also presented in Higgins (2009). Wang et al. state
that, consistent with the findings of previous studies, the
seismic performance of stair systems is highly dependent on
the deformability of their connections.
Figure 1: Damage to stairs observed in full-scale testing at UCSD (from Wang et al. 2015)
There is a growing number of analytical studies which attempt
to quantify the effect of stairs on the building itself. One of the
conclusions of Cosenza et al. (2008) was that in the transverse
building direction, RC stairs contributed to 50 percent of the
building stiffness in that direction. Observed failures of
reinforced concrete stairs in the 2008 Wenchuan and 2010
Yushu earthquakes led to several additional analytical studies.
Xu and Li (2012), Fallahi and Alirezai (2014) and others have
studied the analytical response of RC frame structures with and
without stairs and conclude that the presence of stairs may
significantly alter the dynamic characteristics of a structure
with undesirable consequences. Jiang et. al (2012) reached a
similar conclusion in their study focusing on the potentially
damaging effect of staircases on the building RC frame and
Tegos et al. (2013) studied the response of different stair
configurations including an external staircase on global and
local building behavior.
2. Code Requirements and Project Specifications The design of stairways is governed by codes, which generally
reference ASCE 7 (ASCE, 2010) and is also addressed by
multiple industry publications, design guidelines and
recommendations. The code requirements and breadth of
other information requires significant judgment by the design
professional. Design responsibility is often delegated to an
engineer other than the building structural engineer of record
either through the project construction specifications or, as is
often done, the stairways are defined to be a deferred submittal
item. As a deferred submittal item, their design is not included
in the building structural design developed by the structural
engineer of record, rather it must be developed later, often by
an engineer for a supplier, and submitted in a separate package
for plan check review.
Construction specification documents for stairs vary widely
from project to project. Even in areas of high seismicity the
stair specification may make no mention of seismic
performance. If there is an explicit seismic performance
requirement it is often stated that stair systems should be
capable of withstanding the effects of earthquake motions
determined according to ASCE 7. Some specifications state
that the design of the stairs should be completed by a qualified
engineer while others go so far as to say that the design of the
stairs should include comprehensive engineering analysis by a
qualified professional engineer.
Code requirements for stairs are found in the International
Building Code which in turn makes reference to ASCE 7. The
International Building Code is concerned primarily with the
dimensional requirements for egress stairways such as riser
height, tread width and depth. ASCE 7, Chapter 13 Seismic
Design Requirements for Nonstructural Components requires
that systems that are required for life-safety purposes after an
earthquake, including egress stairways, be classified as
designated seismic systems and be assigned a component
importance factor, Ip, of 1.5. Section 13.3.1 specifies the
seismic design force and Section 13.3.2 requires that the
effects of relative displacements between the ends of the
component be considered. A methodology to calculate the
relative displacement for components with two connection
points is given. In the context of stairs, the relative
displacement given by Eqn. 13.3-6 is the interstory drift from
one floor to the next and includes the building importance
factor, Ie.
2017 SEAOC CONVENTION PROCEEDINGS
3
The guideline document, FEMA E-74, Reducing the Risks of
Nonstructural Earthquake Damage – A Practical Guide
(FEMA, 2015) provides guidance in Section 6.3.8 Stairways
which includes a discussion on typical stair damage and
suggested mitigation techniques. It is recommended that in
order to prevent stairs from acting like diagonal struts between
floors, the design should provide a fixed connection at one
floor and a sliding connection at the other floor that “allows
movement parallel to the direction of the stair”.
Additional guidance for the design of stair systems can be
found in the Metal Stairs Manual (NAAMM, 1992). Section 5
provides engineering data and design examples for metal stair
systems. This manual is often specified under the heading
Quality Control in construction specification documents but
makes no mention of seismic considerations nor developments
since the time of its original publication such as the use of
slotted connections or single end attachments to accommodate
interstory drift.
A stairway design that meets the requirements of ASCE 7
including the explicit requirement that interstory drift be
considered and follows the recommendations of documents
such as FEMA E-74 is a significant challenge for the design
engineer.
3. Full-Scale Dynamic Testing of Stair Systems
Based on the failure of stairs in past earthquakes and the poor
performance observed in recent experimental testing, it is clear
that stairs which are rigidly fixed at the top and bottom landing
are vulnerable in earthquakes and there is an obvious need for
a solution which will accommodate interstory drift within the
stair system. To that end, an experimental testing program was
undertaken as a first step in the development a fully-tested and
engineered solution for stairways.
Shake table testing was conducted at the University of Nevada
Reno on a bi-axial shake table in the Earthquake Engineering
Laboratory. The main objective of the testing was to
demonstrate that flexible stair connections at the top and/or
bottom of the stair assembly can be used to accommodate
building interstory drift. A second objective was to investigate
the performance of the “fixed-free” configuration where the
stairs are fixed at the top and free to move at the bottom
landing, as this configuration has been used and some variant
of it is recommended for new structures in New Zealand via
the Practice Advisory 13 and by FEMA E-74 in the United
States. Finally, the third objective of the testing was to
experimentally evaluate how much force a fixed-fixed stair
system might impart on its surrounding building structure.
3.1 Test Configuration
The test setup comprised a stiff upper landing attached to a
reaction block on the shake table with the lower landing fixed
to the laboratory strong floor. In this configuration, movement
of the table resulted in relative movement between the two
landings simulating interstory drift. The stair specimen itself
was designed so that each end could be replaced allowing
multiple configurations and combinations of top and bottom
connections to be tested using the same central portion of the
stairs. To achieve this, the top and bottom stair and their
connection portions were bolted to the center nine-step section
comprising the stringers, risers and treads. The overall setup
and the stair specimen with removable ends is shown in Figure
2. Many different configurations were tested and six are
described in the results summary of Section 3.4.
The stairs were tested with steel plates attached to the treads to
simulate 100 psf live load, as per ASCE 7-10. In addition,
some configurations were tested with 50 psf live load or no
live load to investigate the effect of the load on the response.
Figure 2: Testing Setup at the University of Nevada, Reno
3.2 Instrumentation Instrumentation included two wire potentiometers at each of
the top and bottom landings to measure longitudinal
displacement and rotation of the stairs as well as one wire
potentiometer to measure the transverse direction at the top
and bottom of the stairs. Tri-axial accelerometers were placed
on the shake table and at the top, mid-height and bottom of the
stairs (Figure 3). Force and displacement of the table are
obtained directly from the shake-table control system.
2017 SEAOC CONVENTION PROCEEDINGS
4
3.3 Loading Protocol The loading protocol included quasi-static, dynamic and
earthquake simulation tests over a range of different
amplitudes. The design interstory drift was assumed to be
2.5% which, for a 10 foot story height, corresponded to 3 in.
of shake table displacement relative to the strong floor. The
MCE drift was assumed to correspond to 4.0% interstory drift
or 4.8 in. of shake table displacement.
Figure 3: Instrumentation Schematic
For each setup configuration, testing was initially conducted
quasi-statically (0.25 in./s) in the longitudinal (X) direction
followed by the transverse (Y) direction to confirm that the
configurations functioned kinematically before conducting
dynamic testing. The tests included three cycles in each
direction at 0.6 in. and 3.0 in. amplitudes. Following the uni-
directional tests, circular tests were conducted, again at 0.6 in.
and 3.0 in. amplitudes. The test sequence was repeated
dynamically (0.5 Hz) with the number of cycles increased to
five at each amplitude. An additional dynamic circular test
with an amplitude of 4.8 in. was conducted to represent the
MCE demand.
In order to investigate the response of the stairs to motion more
representative of earthquake shaking a simulated interstory
drift displacement time-history was developed. A simple
bilinear single degree of freedom building model was assumed
with an elastic period of 0.35 second. The yield strength was
chosen so that there was moderate yielding (ductility of about
5) experienced in both directions for several seconds of
shaking. After investigating the response of a number of bi-
directional earthquake motions, the Newhall record from the
Northridge earthquake was chosen since the response showed
several desirable traits. Namely, the relatively strong shaking
in both directions for 8-10 seconds, the good representation of
cycles with the stairs moving in the same and opposite
directions, several instances when there is a reversal in one
direction without a corresponding reversal in the other and the
fact that the pulse in the Newhall record creates a relatively
strong forward-and-back displacement pulse (Makris and
Black 2003) as shown in Figure 4. A summary of the test
protocol is given in Table 1 below.
Table 1: Summary of Loading Protocol
Direction Amplitude
(in)
Drift
(%) Rate
Quasi-static
Cyclic Tests
(3 cycles)
X 0.6 0.5
0.25
in/s
X 3.0 2.5
Y 0.6 0.5
Y 3.0 2.5
X+Y 0.6 0.5
X+Y 3.0 2.5
Dynamic
Cyclic Tests
(5 cycles)
X 0.6 0.5
0.5 Hz
sine
X 3.0 2.5
Y 0.6 0.5
Y 3.0 2.5
X+Y 0.6 0.5
X+Y 3.0 2.5
X+Y 4.8 4.0
Earthquake
Tests
X+Y Approx.
3” “DBE”
X+Y Approx.
4.8” “MCE”
3.4 Summary of Results
Twelve different mass and end conditions were considered as
part of the test program. The testing looked at three basic
configurations: 1) stairs hanging from the upper landing with
different lower landing configurations; 2) stairs supported on
a track at the upper landing; and 3) stairs fixed at the top
landing and free to move at the lower landing. In addition, the
testing also included a fixed-fixed case, although this proved
to be difficult to test as discussed later.
The following sections provide a brief summary of the
different configurations and the results of the testing. Specific
comparisons between the different configurations are made for
the MCE test. In that test, the landing (shake table) moved
approximately 4 in. in the Y direction and 2.5 in. in the X
direction. This is less than the target of 4.8 in. in the Y direction
at the MCE level. For a number of the configurations the input
signal was increased to 4.8 in. although the comparisons
presented in the following sections are based on the response
to the original MCE input (4 in. maximum landing movement
in Y) as that test was run for every stair system configuration.
2017 SEAOC CONVENTION PROCEEDINGS
5
3.4a Hanger-Spring Configuration
In this configuration, rod end bearings, or heim joints, were
used on either end of a threaded rod to suspend the stairs from
the top landing (Figure 5). These hanger assemblies were
supported by two cantilevers extending from the landing to
provide clearance for the stairs to move towards the landing.
The bottom of the stairs rested on a high-density polyethylene
(HDPE) sliding surface with spring assemblies installed to
provide restoring force in the transverse direction (Figure 6).
This configuration was designed to have unrestricted
movement at the top landing with the bottom connection
allowing rotation and movement with a transverse restoring